A Shoulder to Cry on: Towards A Motivational Virtual Assistant for Assuaging Mental Agony

Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 2436 - 2449 July 10-15, 2022 ©2022 Association for Computational Linguistics A Shoulder to Cry on: Towards A Motivational Virtual Assistant for Assuaging Mental Agony Tulika Saha∗, Saichethan Reddy∗, Anindya Das∗, Sriparna Saha∗, Pushpak Bhattacharyya† ∗Indian Institute of Technology Patna, India †Indian Institute of Technology Bombay, India (sahatulika15,sriparna.saha,pushpakbh)@gmail.com Abstract Mental Health Disorders continue plaguing hu- mans worldwide. Aggravating this situation is the severe shortage of qualified and competent mental health professionals (MHPs), which un- derlines the need for developing Virtual Assis- tants (VAs) that can assist MHPs. The data+ML for automation can come from platforms that allow visiting and posting messages in peer- to-peer anonymous manner for sharing their experiences (frequently stigmatized) and seek- ing support. In this paper, we propose a VA that can act as the first point of contact and comfort for mental health patients. We curate a dataset, Motivational VA: MotiVAte comprising of 7k dyadic conversations collected from a peer-to- peer support platform. The system employs two mechanisms: (i) Mental Illness Classifica- tion: an attention based BERT classifier that outputs the mental disorder category out of the 4 categories, viz., Major Depressive Disorder (MDD), Anxiety, Obsessive Compulsive Dis- order (OCD) and Post-traumatic Stress Disor- der (PTSD), based on the input ongoing dialog between the support seeker and the VA; and (ii) Mental Illness Conditioned Motivational Dialogue Generation (MI-MDG): a sentiment driven Reinforcement Learning (RL) based mo- tivational response generator. The empirical evaluation demonstrates the system capability by way of outperforming several baselines. 1 Introduction With an estimated 970 million individuals suffer- ing from some sort of mental or neural diseases, mental health disorders are regarded one of the pri- mary causes of disability globally1. Poor access, stigma, and prejudice, on the other hand, are likely to limit clinical care to only 15% of individuals who are affected. As a means of expressing their emotions and experiences (generally stigmatized), millions of people (also known as support seekers) 1https://www.singlecare.com/blog/news/mental-health- statistics/ frequently turn to looking for emotional or mental health-related support (Eysenbach et al., 2004) on a variety of text-based peer-to-peer support platforms (De Choudhury and De, 2014), (talklife.co). While peer supports on these platforms are well- intentioned and willing to aid and help seekers, they are often untrained and unaware of best-practices in therapy, resulting in wasted opportunities to provide effective and mutually engaging solutions (Gage-Bouchard et al., 2018). As a result, devel- oping human-computer interfaces in the form of Virtual Assistants (VAs) that can effectively reply and provide support to online support seekers be- comes even more critical. Empathy, or empathetic interactions (Elliott et al., 2018), has been studied extensively in recent years (Sharma et al., 2021, 2020) as one of the most important aspects in providing successful support and triggering beneficial results in support-based dialogues. In addition to empathy, imparting hope and motivation (the process of thinking about and the willingness to move towards one’s goals) …

Content truncated to 500 words